Portfolio Summary

Heather Terada

Where it all started…

  • Willamette University, Bachelor of Arts in Asian Studies (Japanese language) Salem, Oregon, September 2008 - May 2012

  • Master’s of Arts in International Relations Tokyo, Japan, October 2012 - September 2014

From Project Manager to Data Scientist

  • Rakuten Travel, Marketing and Web Design Tokyo, Japan, October 2014 - August 2016

  • Sony Corporation, Risk and Control Department Tokyo, Japan, September 2016 - February 2018

  • MetLife Japan, Application Development Tokyo, Japan, March 2018 - June 2022

  • Sony Corporation of America, Risk and Control Department Remote Oregon, July 2022 - present

New skills in SQL

New skills in R: Tidyverse, ggplot, Plotly, Shiny, Machine Learning

# A tibble: 10 × 6
   bottl earthi herbal ag    berri province  
   <lgl> <lgl>  <lgl>  <lgl> <lgl> <chr>     
 1 TRUE  TRUE   TRUE   FALSE FALSE Oregon    
 2 FALSE FALSE  FALSE  TRUE  TRUE  Oregon    
 3 FALSE FALSE  FALSE  FALSE FALSE California
 4 FALSE FALSE  TRUE   FALSE FALSE Oregon    
 5 FALSE TRUE   FALSE  FALSE TRUE  Oregon    
 6 FALSE FALSE  FALSE  FALSE FALSE Oregon    
 7 TRUE  FALSE  FALSE  FALSE FALSE California
 8 FALSE FALSE  FALSE  FALSE FALSE California
 9 FALSE FALSE  FALSE  FALSE FALSE California
10 FALSE FALSE  FALSE  FALSE FALSE Oregon    
Confusion Matrix and Statistics

                   Reference
Prediction          Burgundy California Casablanca_Valley Marlborough New_York
  Burgundy               207          4                 1           0        0
  California              10        752                20          25       23
  Casablanca_Valley        0          0                 1           0        0
  Marlborough              0          0                 0           3        0
  New_York                 0          0                 0           0        0
  Oregon                  21         35                 4          17        3
                   Reference
Prediction          Oregon
  Burgundy               6
  California            85
  Casablanca_Valley      0
  Marlborough            0
  New_York               0
  Oregon               456

Overall Statistics
                                         
               Accuracy : 0.8482         
                 95% CI : (0.8301, 0.865)
    No Information Rate : 0.4728         
    P-Value [Acc > NIR] : < 2.2e-16      
                                         
                  Kappa : 0.7544         
                                         
 Mcnemar's Test P-Value : NA             

Statistics by Class:

                     Class: Burgundy Class: California Class: Casablanca_Valley
Sensitivity                   0.8697            0.9507                0.0384615
Specificity                   0.9923            0.8152                1.0000000
Pos Pred Value                0.9495            0.8219                1.0000000
Neg Pred Value                0.9787            0.9485                0.9850478
Prevalence                    0.1423            0.4728                0.0155409
Detection Rate                0.1237            0.4495                0.0005977
Detection Prevalence          0.1303            0.5469                0.0005977
Balanced Accuracy             0.9310            0.8829                0.5192308
                     Class: Marlborough Class: New_York Class: Oregon
Sensitivity                    0.066667         0.00000        0.8336
Specificity                    1.000000         1.00000        0.9290
Pos Pred Value                 1.000000             NaN        0.8507
Neg Pred Value                 0.974850         0.98446        0.9200
Prevalence                     0.026898         0.01554        0.3270
Detection Rate                 0.001793         0.00000        0.2726
Detection Prevalence           0.001793         0.00000        0.3204
Balanced Accuracy              0.533333         0.50000        0.8813

Next steps

  • Applying new skills to current work

  • Future roles in Data Science